Earlier this month Imagination Technologies picked up two awards for its recently launch IMG Series 4 neural network accelerator (NNA). Our NNA is no stranger to awards, with the design team behind the Series2 NX and the NNA itself picking up gongs for its ground-breaking performance.
The Series 4 takes things to new levels, with its scalable architecture offering 500 TOPS of performance, and even beyond. This makes it ideal for the automotive market, and next-generation ADAS and full self-driving systems, and we already have licensees in that market.
It’s pleasing that the Series4 has already been recognised for its strength in these areas and at the World Smart Car Conference held in Guangzhou, China on 4 December 2020, Imagination picked up the “Autonomous Driving Award”. At the event, Liu Guojun, Vice President and General Manager of Imagination China delivered a speech on “Globalization, Technological Innovation and Change”.
On the very same day, Imagination also collected the “Highest Potential Enterprise Award” at the China IoT Conference and the 5th China IoT Innovation Awards in Shenzhen.
Imagination doesn’t just turn up to win awards in China: it also provides insight and knowledge. At the “IoT Industry Development Summit Forum”, Huang Yin, Senior Business Development Manager of Imagination, delivered a keynote speech on how “SoC IP Helps Realise the Smart Internet of Things.”
In her speech, Huang Yin demonstrated some AI applications based on Imagination NNA technology, such as face detection and recognition, gesture detection, mouth and eye-opening/closing detection.
In addition, Huang Yin also pointed out that Imagination’s NNA+GPU can form a heterogeneous platform with both graphics processing and AI computing functions, and that it has been successfully applied in personnel detection, driver monitoring and car surround view.
Imagination also appeared at the GTIC 2020 AI Chip Innovation Summit, where Vice President of Imagination and General Manager of China Region delivered a keynote speech on how multi-core GPUs and dedicated NNA can promote intelligent applications from the Cloud to the Edge.